Legal Beagle by Graeme Edgeler

Read Post

Legal Beagle: At least we have MMP

101 Responses

First ←Older Page 1 2 3 4 5 Newer→ Last

  • Graeme Edgeler,

    There's no difference, overhang is tactical advantage. Both increase the chance of a voter getting their representatives into government.

    Sure, but what I was looking at was whether the Maori seats are inherently more like to cause an overhang. They are.

    That this *is* a tactical advantage does not mean it is caused by tactical behaviour. My analysis leads me to the conclusion that it is not behaviour from Maori Party voters, or Maori roll voters that primarily causes this, but the relative youth of the Maori descent population, coupled with lower enrolment and turnout.

    Wellington, New Zealand • Since Nov 2006 • 3215 posts Report Reply

  • Rich of Observationz,

    NZF won all five Maori seats in 1996 and lost them on 1999 after fallouts and waka jumping.

    I think various National MPs owe their seats to the Maori roll, which limits the working class vote in much of the rural North Island.

    Back in Wellington • Since Nov 2006 • 5550 posts Report Reply

  • Kyle Matthews,

    That this *is* a tactical advantage does not mean it is caused by tactical behaviour. My analysis leads me to the conclusion that it is not behaviour from Maori Party voters, or Maori roll voters that primarily causes this, but the relative youth of the Maori descent population, coupled with lower enrolment and turnout.

    If Maori were (as Craig calls it) Te Borg, the tactical vote implications would be quite impressive. Up to 13 electoral seats all overhung, with all the party votes going to a party which would enter into a pre-arranged coalition with the Maori Party. Balance of power almost guaranteed.

    Sadly, people don't work like that.

    Since Nov 2006 • 6243 posts Report Reply

  • Mikaere Curtis,

    That's just how the Sainte-Laguë method works out, which I note isn't applied in his analysis.

    Yeah, forgot about that. It's not easy to apply, so I'll code up an application to do it tonight and repost.

    I suspect that the main difference between my and your numbers will be in the assumption of voter turnout rates. My assumption is larger seats, same rates; I predict a 1 seat overhang. More tonight...

    Tamaki Makaurau • Since Nov 2006 • 528 posts Report Reply

  • uroskin,

    It's akin to the idea that voting for anyone but National or ACT was pointless in the last election because they won, so everyone else got no power.

    We don't vote for power but for representation. In the current MMP system with Maori seats and thresholds, the representation is distorted by electorate seats, their possible overhang and the coat tail MPs who get in because one portion of their electorate is so concentrated or gerrymandered their vote is worth far more than a party that gets 4.5% but spread nationally.
    Any review of MMP should make it more proportional by ridding us of Maori and general seats.

    Waiheke Island • Since Feb 2007 • 178 posts Report Reply

  • Angus Robertson,

    Mikaere's numbers are wrong (my apologies for not noticing his comment earlier). He assumes that you earn a seat with c. 20k votes. This is false. Without a threshold, at the last election, a party could have earned a seat with 9,160 votes. Three seats would have been earned with 46,611 votes, and five seats with 85,411 votes. That's just how the Sainte-Laguë method works out, which I note isn't applied in his analysis.

    Actually I think we do have a 5% threshold.

    That this *is* a tactical advantage does not mean it is caused by tactical behaviour. My analysis leads me to the conclusion that it is not behaviour from Maori Party voters, or Maori roll voters that primarily causes this, but the relative youth of the Maori descent population, coupled with lower enrolment and turnout.

    They got less than 5% of the party vote and didn't win the party vote in any single electorate. Since Graeme has concluded that tactical voting was insignificant, I am really interested to see how they get them into parliament under any extrapolated bias elimination process. How big are these electorates going to be?

    Auckland • Since May 2007 • 984 posts Report Reply

  • Graeme Edgeler,

    Actually I think we do have a 5% threshold.

    Yes. But because the Maori Party won an electorate it doesn't matter for them. Had they earned from 0 - 9,159 votes they'd have caused an overhang of 5 seats. Had they earned 85,411+ votes they'd have caused no overhang.

    Wellington, New Zealand • Since Nov 2006 • 3215 posts Report Reply

  • BenWilson,

    We don't vote for power but for representation.

    I can't speak for you, but I vote in the hope that the representatives I select might have some influence on power. Since the party I voted for has no influence on the ability of the legislature to pass laws, and holds no ministerial positions I lucked out. Ergo, my vote was worthless.

    I hope you can see I'm NOT actually taking this position, but showing you that your reductio ad absurdum can go both ways. You drove to the ridiculous conclusion that Maori party votes were infinitely more valuable than NZF votes, based on the outcome. I'm showing it extends in the opposite direction. Basically, I'm saying that the relationship isn't the simple (outcomes/votes) ratio you're suggesting. It's much more complicated than that.

    But I do agree with you, I think, on the whole, that the fact that the outcomes are far from equal is not ideal. I'd like to push towards ideal, whilst understanding that there's still going to be imperfection. Some of the imperfection even has strong arguments, like the thresholds and the Maori seats. I just happen not to agree with them.

    Auckland • Since Nov 2006 • 10657 posts Report Reply

  • Angus Robertson,

    Yes. But because the Maori Party won an electorate it doesn't matter for them.

    How did they win 5 electoral seats?

    I believe they were the beneficiaries of tactical voting.

    Had they earned 85,411+ votes they'd have caused no overhang.

    And they wouldn't have any seats, unless there was a significant amount of tactical voting.

    Auckland • Since May 2007 • 984 posts Report Reply

  • Tim McKenzie,

    If Maori were (as Craig calls it) Te Borg, the tactical vote implications would be quite impressive.

    Actually, the Māori Party could do some quite spectacular tactical manipulations without requiring any tactical voting. Consider this strategy:

    Tariana Turia stands in Te Tai Hauāuru as normal. All other Māori Party electorate candidates who have a reasonable chance of winning (probably just the ones in Māori electorates) resign from the Māori Party and stand as "independents". The Māori Party and the "independent" candidates make it plain that they still support each other as much as ever.

    Then, based on the last election, something like this happens:

    Tariana Turia and four supportive "independents" win their electorates. Because there are four independents, the proportionality of Parliament is worked out for the remaining 116 seats. The Māori Party gets its fair proportion of those 116 seats (coming in on Tariana Turia's coat-tails). But in addition to that fair proportion, they also get the four "independent" candidates, giving them a much higher proportion of Parliament than their party votes entitled them to.

    Any party could benefit by using this strategy, but it works best for parties like the Māori Party that win disproportionately many electorate seats compared to their party vote. If all parties did it, you'd effectively end up with Supplementary Member.

    So why doesn't anyone try this? I suppose the risk for parties that get disproportionately few electorates compared to their party vote is that they'll give other parties ideas, and end up suffering from de facto Supplementary Member. But if someone like the Māori Party starts it, they should seriously consider joining in, to mitigate their suffering.

    Lower Hutt • Since Apr 2007 • 126 posts Report Reply

  • Tim McKenzie,

    The voters in an average Maori electorate have enough party votes between them to elect 1.051 MPs. The voters in the average general electorate (because there are many more of them per electorate) have enough party votes between them to elect 1.849 MPs.

    Thinking about this again, I've come to the conclusion that my objection is this:

    When I first read it, I got the impression (whether you intended it or not) that you were saying that Māori electorates were disadvantaged, because they only elected 1.051 MPs each, compared to 1.849 MPs per general electorate. Therefore (so I thought) you were suggesting that voters in the "disadvantaged" electorates were also disadvantaged. As discussed above, this is not the case; voters in "disadvantaged" electorates have more influence than the average voter on the distribution of electorate seats, and if this extra influence results in an overhang (as you argue is very likely), then they have more influence on the overall distribution of seats in Parliament

    Lower Hutt • Since Apr 2007 • 126 posts Report Reply

  • Mikaere Curtis,

    I've revised my numbers as if the Maori electorates were increased in size to the same as the average general electorates, and maintained the same voter turnout and voting patterns.

    Accordingly, the adjusted party votes are therefore:

    National 1056902.312
    Labour 821328.431
    Greens 159476.183
    ACT 85726.979
    Maori 70145.403
    Progressives 21381.301
    United Future 20583.394

    Then I wrote a program to calculate the Sainte-Laguë results, and they are:

    National 56
    Labour 44
    Greens 9
    ACT 5
    Maori 4
    United Future 1
    Progressives 1

    So, there is still an overhang. Unless you use Graeme's assumptions and change not only the size, but the voting turnout as well.

    Tamaki Makaurau • Since Nov 2006 • 528 posts Report Reply

  • Graeme Edgeler,

    I got the impression (whether you intended it or not) that you were saying that Māori electorates were disadvantaged, because they only elected 1.051 MPs each, compared to 1.849 MPs per general electorate. Therefore (so I thought) you were suggesting that voters in the "disadvantaged" electorates were also disadvantaged. As discussed above, this is not the case; voters in "disadvantaged" electorates have more influence than the average voter on the distribution of electorate seats...

    The voters have more influence, but the electorate has less. Electorates are all roughly equally sized, representing approximately 60,000 people each. The 60,000 people represented in your average Maori electorate have far less influence over an election result than the 60,000 people represented in your average general electorate - even taking into account overhang.

    Why is this? First, because far fewer of them vote, but most importantly because far fewer of them can vote. And when you're working out influence over the election result it's raw numbers in the party vote that matter most.

    Wellington, New Zealand • Since Nov 2006 • 3215 posts Report Reply

  • Graeme Edgeler,

    How did they win 5 electoral seats?

    I believe they were the beneficiaries of tactical voting.

    That's one way to look at it.

    You're looking at the results as showing that the Maori Party were the beneficiaries of tactical voting in the Maori seats.

    Another way to look at it is that the Labour Party were the beneficiaries of tactical voting in the Maori seats: Maori Party supporters looked at the polls, realised that the Maori Party were unlikely to win any list seats, so party voted for the Labour Party, giving them some extra seats.

    I don't think either of these is the correct way or the incorrect way to look at the results, but is does show how our biases can colour our interpretation of data.

    Wellington, New Zealand • Since Nov 2006 • 3215 posts Report Reply

  • Angus Robertson,

    You're looking at the results as showing that the Maori Party were the beneficiaries of tactical voting in the Maori seats.

    Yeah and pre-election polling agrees, but polls tend to focus solely on party vote so are inaccurate in regard to tactical voting.

    Another way to look at it is that the Labour Party were the beneficiaries of tactical voting in the Maori seats: Maori Party supporters looked at the polls, realised that the Maori Party were unlikely to win any list seats, so party voted for the Labour Party, giving them some extra seats.

    A marvellous description of how they could vote tactically to ensure an overhang.

    I don't think either of these is the correct way or the incorrect way to look at the results, but is does show how our biases can colour our interpretation of data.

    In Epsom my biases lead me to suspect lots in that electorate voted tactically, because it was a smart thing to do. Do you agree?

    In the Maori electorates my biases lead me to suspect lots in those electorates voted tactically, because it was a smart thing to do. Do you agree?

    Auckland • Since May 2007 • 984 posts Report Reply

  • Graeme Edgeler,

    I'm not disputing that tactical voting can cause an overhang, nor that it helped cause one here. I'm suggesting the larger cause of overhang is an unequal distribution of voters between electorates.

    And mostly, I'm backing up my riposte to your assertion that the Maori electorates are not inherently more likely to cause an overhang. Whether you agree that this overhang was strongly correlated to voter numbers across the Maori electorates or not, I think my argument at the very least shows that overhang is more likely in electorates with lower than average voter numbers (which the Maori electorates are).

    Wellington, New Zealand • Since Nov 2006 • 3215 posts Report Reply

  • uroskin,

    What a truly proportional parliament would have looked like in 2008 (no electorate seats, 0.8% threshold):
    National: 44.93% = 53.92 seats = 54 (-4)
    Labour: 33.99% = 40.80 seats = 41 (-2)
    Green: 6.72% = 8.06 seats = 8 (-1)
    NZ First: 4.07% = 4.88 seats = 5 (+5)
    ACT: 3.65% = 4.38 seats = 4 (-1)
    Maori Party: 2.39% = 2.87 seats = 3 (-2)
    Progressive: 0.91% = 1.09 seats = 1 (-)
    United Future: 0.87% = 1.04 seats = 1 (-)

    A "Maniact" Government would still have been a possible outcome.

    Waiheke Island • Since Feb 2007 • 178 posts Report Reply

  • Angus Robertson,

    I'm not disputing that tactical voting can cause an overhang, nor that it helped cause one here. I'm suggesting the larger cause of overhang is an unequal distribution of voters between electorates.

    The numbers, they don't work out.

    2008 Maori electorates produced 33,807 party votes for the Maori Party. To increase this to 85,000+ (ensuring they have 5 PR MPs) requires making each electorate 2.5x larger than they currently are. Each electorate would then have 52,000 active voters or 40% more than the general roll average.

    This isn't equivalency.

    And mostly, I'm backing up my riposte to your assertion that the Maori electorates are not inherently more likely to cause an overhang. Whether you agree that this overhang was strongly correlated to voter numbers across the Maori electorates or not, I think my argument at the very least shows that overhang is more likely in electorates with lower than average voter numbers...

    A situation that, whilst being theoretically logical, doesn't apply in NZ.

    ...(which the Maori electorates are).

    Not under MMP, at least not for party vote. The party vote that determines the proportionality of parliament is counted across all electorates. Each electorate MP is effectively drawing upon an equal and very large electorate to determine their party vote.

    Auckland • Since May 2007 • 984 posts Report Reply

  • Graeme Edgeler,

    2008 Maori electorates produced 33,807 party votes for the Maori Party. To increase this to 85,000+ (ensuring they have 5 PR MPs) requires making each electorate 2.5x larger than they currently are.

    Because we assume the Maori Party wouldn't get 22,000+ party votes in the general electorates?

    Wellington, New Zealand • Since Nov 2006 • 3215 posts Report Reply

  • Angus Robertson,

    Because we assume the Maori Party wouldn't get 22,000+ party votes in the general electorates?

    You appear to be saying so by...

    ...suggesting the larger cause of overhang is an unequal distribution of voters between electorates.

    To state that causation is due to unequal distribution of voters, whilst basing your calculations on a tally derived by equally distributing voters from all the electorates is not going to be valid.

    33,807 people were the Maori Party vote in the smaller and unequal Maori seats. 55,980 people were the Maori Party vote in the whole of NZ, an electorate that is definitely not small or unequal.

    Auckland • Since May 2007 • 984 posts Report Reply

  • Tim McKenzie,

    The 60,000 people represented in your average Maori electorate have far less influence over an election result than the 60,000 people represented in your average general electorate - even taking into account overhang.

    OK, I see your point. (Assuming your 1.051 MPs per Māori electorate is roughly accurate.)

    Why is this? First, because far fewer of them vote, but most importantly because far fewer of them can vote.

    These two factors can be addressed separately, and I'd be interested to know how much each contributes.

    There are several ways you could address the first issue:
    1. It's not a problem; it's perfectly valid to choose not to cast a party vote, letting the rest of the country decide the proportionality of Parliament.
    2. It is a problem, but it should be addressed culturally, by encouraging people to vote (or discouraging people in other electorates), rather than by changing the system.
    3. Make it compulsory to vote.
    4. Weight party votes so that all eligible voters in an electorate are assumed to have cast their party votes in proportion to how party votes were actually cast by those who actually used their party votes in that electorate; this effectively assumes that people who choose not to use their party vote should be represented by people in their electorate who did use it, rather than by people across the country who used it.

    There are several ways you could address the second issue:
    1. It's not a problem; Parliament's proportionality should be determined by numbers of voters who prefer each party, not by the sizes of their families.
    2. It is a problem, but trying to fix it would break something else.
    3. Give children the vote (and their parents the right to cast it for them, up to a certain age).
    4. Weight party votes according to the ratio of (total population):(eligible voters) in an electorate; this effectively assumes that children are best represented by voters in their electorate, rather than by voters across the country.

    If I've understood Mikaere's analysis correctly, he effectively took option 4 from the second list (perhaps with the simplifying assumption that the relevant ratio was one constant across all Māori electorates and another constant across all general electorates). But there was still an overhang, distorting the proportionality in Parliament.

    So, if you agree with option 1 from both lists, you should be upset that there's an overhang distorting the proportionality of Parliament. If you agree with option 1 from the first list and option 4 from the second list, Mikaere's analysis shows that there would still be an overhang, still distorting the proportionality of Parliament. But if you agree with option 4 from both lists (and assuming the 1.051 MPs per Māori electorate figure is roughly correct), then you should be upset that the Māori electorates are still under-represented in Parliament, in spite of the overhang.

    Lower Hutt • Since Apr 2007 • 126 posts Report Reply

  • jeremy gray,

    Tim, your argument about splitting a party into two parties contesting electorate and list votes happened in Italy in 2001. The two main parties put all their electorate members in a decoy party and asked for voters to not give their paty votes to the original party. The problem was both sides did the same, and it kind of evened out.

    point chev • Since Apr 2008 • 44 posts Report Reply

  • Tim McKenzie,

    Tim, your argument about splitting a party into two parties contesting electorate and list votes happened in Italy in 2001.

    Interesting. Thanks for pointing it out.

    I was suggesting standing nominal independents, rather than putting electorate candidates in another party. For the Māori Party, this mightn't make much difference, but if the National Party (and only the National Party) had done it at the last election, I think standing nominal independents would have got them about 66% of Parliament, compared to "only" about 61% (of a larger Parliament) if they'd used two parties.

    If all parties did it, I think you'd end up with de facto Supplementary Member, which (in comparison to MMP) favours parties (like the Māori Party) more capable of winning electorates.

    This is all assuming that my understanding of how MMP handles independents is correct.

    Lower Hutt • Since Apr 2007 • 126 posts Report Reply

  • Graeme Edgeler,

    This is all assuming that my understanding of how MMP handles independents is correct.

    Tim - it appears to be. It is something that a lot of people get wrong. If anyone who does not come from a party that is contesting the party vote wins an electorate, proportionality is maintained over less than 120 seats. Independents, and candidates from parties (registered or not) that do not submit party lists, do not cause overhang.

    Wellington, New Zealand • Since Nov 2006 • 3215 posts Report Reply

  • Angus Robertson,

    Because we assume the Maori Party wouldn't get 22,000+ party votes in the general electorates?

    When correcting for any disparity caused by the smaller size of the Maori electorates, bringing in more voters in equivalent proportions to how the electorates did vote, I assumed you were going to be bringing them in from the general roll electorates. Mostly I assumed this because these voters actually exist, they are real.

    I suspect however that you were suggesting about 100,000 more Maori voters would magically pop into existance so the Maori Party gets 30,000 more votes. That is simply making shit up.

    Auckland • Since May 2007 • 984 posts Report Reply

First ←Older Page 1 2 3 4 5 Newer→ Last

Post your response…

Please sign in using your Public Address credentials…

Login

You may also create an account or retrieve your password.